alcohol preference in self-administration models (Li et al., 1993). These animal models have characteristics that mirror in many respects human alcohol addiction; the animals show tolerance development, and pharmacological agents (e.g., naltrexone, an opiate antagonist) that reduce alcohol consumption in humans also reduce consumption in the animals (Li and McBride, 1995; Schuckit, 1994). Inbred and recombinant strains of mice have also been developed to assess the contributions of multiple genes to alcohol- (and drug-) related responses. Using a method called quantitative trait loci gene mapping (QTL), specific parts of genes have been identified that are associated with a variety of alcohol-related traits, including alcohol preference, sensitivity, tolerance, and withdrawal (Crabbe et al., 1994).
The search for specific genes that may confer enhanced risk of alcoholism has already progressed significantly, and much interest is focused on variant genes that may increase vulnerability to other types of addiction. The study of environmental factors that produce risk versus resilience with respect to drug addiction, however, must contend with multiple behavioral, psychological, and social factors that complicate analysis. In addition, other factors influence drug taking, including drug availability, acceptability in the individual's subculture, behavioral alternatives to drug use, and presence of other psychiatric illness. Factors that influence the "capture rate" include the intrinsic addictiveness of the drug used and, as mentioned above, the form in which it is administered (e.g., smoked crack cocaine).
In the model that has been outlined here, addiction is a disease of the nervous system, interacting with other biological systems and behavioral and environmental factors, that markedly impairs a person's ability to control his or her drug-seeking behavior; in fact, it makes resumption of drinking or drug-taking a compelling risk for the individual despite the highly negative effects of drug addiction on the individual's health and well-being.
Without building a greater context, however, such a biological model could contribute to the common misconception that if alcoholism or drug addiction is not simply willful misbehavior, addicted individuals cannot be held responsible for their actions. Individuals predisposed to developing coronary artery disease, for example, are not castigated by their physicians, but they are asked to follow a certain diet, to exercise, and to comply with medication regimens. Yet, for many individuals, the behavioral changes required to cope with heart disease are difficult and relapses to well-ingrained patterns are common. The same concept should be applied by physicians and society to those addicted to alcohol, nicotine, or illicit drugs. As in the treatment of heart disease, the best therapeutic results are obtained when addicted persons are given responsibility for themselves once a diagnosis is made and treatment recommended.
The clinically critical difference between addiction and coronary artery disease, however, is that the disease of addiction markedly diminishes the ability of the patient to follow through on medical advice. Hard as it is for patients with coronary artery disease to comply with treatment it is even harder for addicted